The proposed study intends to investigate the effects of four parametrically determined drug dosages of methylphenidate (Ritalin), administered in mg/kg, on recall memory in a sample of male children diagnosed as hyperactive. Through the use of a paired-associate learning task, our first objective is to establish dose-response curves for both the accuracy of responding and trials to criterion in order to determine optimal dosage for the acquisition and retention of learned material. Our second objective is to determine if state-dependent learning occurs under this stimulant medication for any of the doses used in this study. Our design is a three factor ANOVA with repeated measures (drug dosage, retention state, and acquisition state) employing a 2 x 2 placebo drug crossover for the last two factors. On our task, we hypothesize that an intermediate dose of 0.3mg/kg will show results superior to placebo, while the effects of a lower dose (0.lmg/kg) will be suboptimal. Higher doses (0.5mg/kg and 1.0mg/kg) will decrease learning and demonstrate state-dependent learning effects when retention is measured in a different state, i.e., placebo. We suggest that individually titrated drug dosage, based on clinical reports, doses not give sufficient control over the independent variable in order to determine the effect of the drug on cognitive behavior. The establishment of drug dose-response curves and an examination of state-dependent learning will help us determine if cognitive decrements or memory dissociation occur at doses which are currently recommended for the control of overactive social behavior.